CN111464231B - Unmanned aerial vehicle and user cooperative cache placement method and device - Google Patents

Unmanned aerial vehicle and user cooperative cache placement method and device Download PDF

Info

Publication number
CN111464231B
CN111464231B CN202010255135.8A CN202010255135A CN111464231B CN 111464231 B CN111464231 B CN 111464231B CN 202010255135 A CN202010255135 A CN 202010255135A CN 111464231 B CN111464231 B CN 111464231B
Authority
CN
China
Prior art keywords
user
unmanned aerial
cache
aerial vehicle
iteration
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010255135.8A
Other languages
Chinese (zh)
Other versions
CN111464231A (en
Inventor
张天魁
陈超
许文俊
曾志民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN202010255135.8A priority Critical patent/CN111464231B/en
Publication of CN111464231A publication Critical patent/CN111464231A/en
Application granted granted Critical
Publication of CN111464231B publication Critical patent/CN111464231B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/14—Relay systems
    • H04B7/15—Active relay systems
    • H04B7/185—Space-based or airborne stations; Stations for satellite systems
    • H04B7/18502—Airborne stations
    • H04B7/18506—Communications with or from aircraft, i.e. aeronautical mobile service
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/70—Services for machine-to-machine communication [M2M] or machine type communication [MTC]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Astronomy & Astrophysics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

The invention provides a method and a device for cooperatively placing caches of an unmanned aerial vehicle and a user, wherein the method comprises the following steps: obtaining cache placement information of the unmanned aerial vehicle by an alternating direction multiplier method, and iterating the cache placement information of the user by a global greedy algorithm; continuously iterating the unmanned aerial vehicle cache placement and the user cache placement, and judging whether the maximum iteration number is reached; if the iteration does not reach the maximum number of times, adding 1 to the iteration number, and iterating the cache placement information of the unmanned aerial vehicle and the cache placement information of the user again; if the iteration reaches the maximum times, outputting the cache placement information of the unmanned aerial vehicle and the user at the moment; the technical scheme of the embodiment of the invention can solve the technical problem that the cache placement decision of the unmanned aerial vehicle with the cache capability and the user is not optimized in a combined manner in the prior art.

Description

Unmanned aerial vehicle and user cooperative cache placement method and device
Technical Field
The invention relates to the technical field of unmanned aerial vehicle communication technology, in particular to a method and a device for cooperatively caching and placing an unmanned aerial vehicle and a user.
Background
With the rapid development of small unmanned aerial vehicles supporting wireless communication, unmanned aerial vehicles have been gradually applied to practical scenes such as disaster recovery, hot spot coverage, relay communication, information dissemination/data collection, and become one of effective technologies for solving certain specific communication scenes. In a cell with dense users, massive data requests bring huge pressure to base station loads, and the unmanned aerial vehicle can be rapidly deployed around the users as an aerial base station by virtue of good flight characteristics of the unmanned aerial vehicle, assists a cellular network to provide communication services for ground users, expands system capacity and improves user service quality. The repeated transmission of a few popular contents is one of typical characteristics of internet content distribution, the request and transmission of the popular contents are more intensive in a cell of a high-density user, and the edge cache is used as an effective technology, so that the contents are cached in advance on an edge node closer to the user and are directly transmitted to the user during peak flow, network congestion can be effectively reduced, and meanwhile, the time delay for the user to acquire the contents is reduced.
In the unmanned aerial vehicle communication network, the unmanned aerial vehicle can carry the storage device to carry out content caching, so that not only can a user acquire content more quickly, but also the consumption of the limited bandwidth resource of the wireless backhaul link of the unmanned aerial vehicle can be reduced. However, the size and the load weight limit the buffer space of the drone, and in order to expand the storage capacity of the network and improve the buffer performance of the network, the storage resources of the user equipment can be used for content buffering, and the content can be shared and spread among users by combining the technology of communication (D2D) among the devices. On one hand, the content requested by the user can be acquired from other users at a closer distance through the D2D link, and the time delay can be further reduced; on the other hand, repeated data transmission between the unmanned aerial vehicle and the user is not required to be frequently performed for many times, and energy consumption caused by downlink transmission of the unmanned aerial vehicle is effectively reduced. However, in the face of an extremely large data volume in a network, how to cache the most appropriate content in an extremely limited storage space is an important problem to be solved, and especially when the unmanned aerial vehicle and the user equipment have the caching capability at the same time, how to cooperate with the caching to reduce the content acquisition delay of the user needs to design an effective cache placement strategy.
For an unmanned aerial vehicle communication network introducing a cache technology, the work of predecessors does not consider the scene that both the unmanned aerial vehicle and a user have cache capacity and the design of a cache placement strategy of the unmanned aerial vehicle and the user at the moment. When the unmanned aerial vehicle and the user are provided with caches, because the unmanned aerial vehicle and the user are located at different positions in the network, the contents cached by the unmanned aerial vehicle and the user are not only related but also differentiated, and how to respectively determine the contents cached by the unmanned aerial vehicle and the user makes the performance of the whole network maximized is very important. Therefore, the cache placement of the unmanned aerial vehicle and the user equipment is optimized in a combined mode, the content acquisition time delay of the user is reduced by designing an effective unmanned aerial vehicle and user cooperative cache strategy, and the system performance is improved.
Disclosure of Invention
The invention aims to provide a method and a device for cooperatively caching and placing an unmanned aerial vehicle and a user in an unmanned aerial vehicle communication network, which solve the technical problem that the cache placing decision of the unmanned aerial vehicle with the cache capability and the user is not optimized in a combined manner in the prior art.
In order to achieve the above object, the present invention provides a method for cooperatively caching and placing an unmanned aerial vehicle and a user in an unmanned aerial vehicle communication network, which comprises the following steps:
obtaining cache placement information of the unmanned aerial vehicle by an alternating direction multiplier method, and obtaining user cache placement information by a global greedy algorithm;
continuously iterating the cache placement information of the unmanned aerial vehicle and the cache placement information of the user, and judging whether the maximum iteration times is reached;
if the iteration does not reach the maximum number of times, adding 1 to the iteration number, and iterating the cache placement information of the unmanned aerial vehicle and the cache placement information of the user again;
and if the iteration reaches the maximum times, outputting the information of the unmanned aerial vehicle and the user cache placement at the moment.
The method for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network comprises the following steps of:
step 201: initializing cache placement information of the unmanned aerial vehicle k in a random mode;
step 202: and (4) updating each cache variable x of the unmanned aerial vehicle k according to the formula (1) by using a gradient descent algorithm according to any iteration number tt k,mTo obtain xt+1 k,m：
Figure BDA0002437006550000031
Wherein L is an augmented Lagrange function of the content acquisition delay D of the whole network user,
Figure BDA0002437006550000032
is an arbitrary unmanned aerial vehicle,
Figure BDA0002437006550000033
is a set of unmanned aerial vehicles,
Figure BDA0002437006550000034
in order for any user to be able to do so,
Figure BDA0002437006550000035
in order to be a set of users,
Figure BDA0002437006550000036
in the case of any content, the content is,
Figure BDA0002437006550000037
is a collection of contents, xk,mCaching variables for unmanned aerial vehicles, xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0, matrix
Figure BDA0002437006550000038
Express all unmanned aerial vehicle's buffer memory and place information, unmanned aerial vehicle k's buffer memory is Q for spacekIndividual content representation, λkLagrange multiplier for drone k;
step 203: all buffer variables x are expressed according to equation (2)t+1 k,mThe value is fixed between 0 and 1:
Figure BDA0002437006550000039
step 204: according to the formula (3), updating the Lagrange multiplier lambda corresponding to the unmanned aerial vehicle kk：λt kTo obtain lambdat+1 k：
Figure BDA00024370065500000310
Rho is a constant parameter for adjusting the convergence speed of the alternative direction multiplier method;
step 205: sequentially updating the cache variable, fixing the range of the cache variable and updating the Lagrange multiplier, and continuously iterating to judge whether the user content acquisition delay value after each iteration period is not changed any more than the last iteration period, namely | D (X)t+1 k)-D(Xt k)|<ε, ε is a small constant parameter, where D (X)t+1 k) Representing the user content acquisition delay value, D (X), after the current iteration cyclet k) Representing the time delay value of the previous iteration period; if the user delay D is changing, go to step 206; otherwise, go to step 207;
step 206: adding 1 to the iteration times;
step 207: and (4) finishing iteration, and performing relaxation removal under the limitation of the cache space of the unmanned aerial vehicle to be processedThe value of the cache variable between 0 and 1 is converted into 0 or 1, and the optimal cache content X of the unmanned aerial vehicle k is obtainedk。
In the method for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network, L is an augmented Lagrangian function of the content acquisition delay D of the users in the whole network, and D is the content acquisition delay of the users in the whole network:
Figure BDA0002437006550000041
wherein λ iskBeing the lagrangian multiplier for drone k,
Figure BDA0002437006550000042
for the penalty term, ρ is a constant parameter for adjusting the convergence rate of the alternative direction multiplier method, and fn,mRepresenting the preference of user n for content m, S representing the size of the content, b being a ground base station, Rn,n′,Rk,n,Rb,nRespectively, the transmission rate of the wireless link between the user n and the adjacent user n', the unmanned aerial vehicle k and the base station B, Bk,nThe wireless backhaul link transmission rate assigned to user n for drone k,
Figure BDA0002437006550000043
for user access indication u n,n′1 means that user n establishes a D2D communication link with a neighboring user n' to obtain content, whereas un,n′＝0，uk,n1 means that the user n accesses the unmanned plane k to acquire the content, otherwise uk,n＝0，ub,n1 indicates that user n accesses to the ground base station b to obtain the content, otherwise ub,n＝0，yn',mCaching variables for users, y n',m1 denotes that user n' caches content m, whereas yn',m＝0。
The method for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network comprises the following steps in step 202:
step 301: calculating the cache variable x of the augmented Lagrange function under the condition of any gradient descent iteration times pp k,mGradient (2):
Figure BDA0002437006550000051
wherein
Figure BDA0002437006550000052
λkIn order to be a lagrange multiplier,
Figure BDA0002437006550000053
is a cache variable;
step 302: updating the cache variable according to equation (4):
Figure BDA0002437006550000054
wherein p is iteration times, and l is learning rate;
step 303: update the learning rate lp+1＝lp(0.96p)；
Step 304: continuously iterating the three steps of calculating gradient, updating cache variable and updating learning rate, and judging whether the value of the function L after each iteration period is not changed any more than the last iteration period, namely | L (x)p+1 k,m)-L(xp k,m)|<Epsilon, epsilon is a small constant parameter; if the lagrangian function value L is changing, go to step 305; otherwise, go to step 306;
step 305: adding 1 to the iteration times;
step 306: after the iteration is finished, outputting the currently updated cache variable xt+1 k,mThe value of (c).
The method for cooperatively placing the cache of the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network comprises the following steps of obtaining user cache placing information through a global greedy algorithm:
step 401: in any iteration cycle s of the greedy algorithm, calculating the content acquisition time delay delta D (delta D) of all network users, which can reduce the content of each user cache, and D (y)n,m＝1)-D(yn,m0), wherein, yn,mCaching variables for users, y n,m1 denotes that user n caches content m, whereas yn,m0, matrix
Figure BDA0002437006550000055
Cache placement information representing all users;
step 402: and comparing and finding out a user content pair with the largest delay reduction degree: { n, m } ═ max Δ D;
step 403: having the user store the content: y isn,m＝1；
Step 404: the above steps are iterated continuously, and whether the cache spaces of all the users are full after each iteration period is judged, namely that
Figure BDA0002437006550000061
Wherein QnRepresenting a cache space of user n; if yes, go to step 406; otherwise, go to step 405;
step 405: adding 1 to the iteration times;
step 406: and (5) finishing the iteration and outputting the optimal user cache placement information Y.
The utility model provides a device that unmanned aerial vehicle and user cooperate buffer memory to place in unmanned aerial vehicle communication network, the device deploys on macro base station, includes: an unmanned aerial vehicle optimal cache placement processor, a user optimal cache placement processor and an optimization control processor; wherein,
the optimal cache placement processor of the unmanned aerial vehicle is connected with the optimal cache placement processor of the user and the optimization control processor, and the optimal cache placement information of the unmanned aerial vehicle at the moment is obtained by utilizing an alternating direction multiplier method according to the user cache placement information input by the optimization control processor and is input into the optimal cache placement processor of the user;
the user optimal cache placement processor is connected with the unmanned aerial vehicle optimal cache placement processor and the optimization control processor, and according to unmanned aerial vehicle cache placement information input by the unmanned aerial vehicle optimal cache placement processor, the optimal user cache placement information at the moment is obtained by using a global greedy algorithm and is input into the optimization control processor;
the optimization control processor is connected with the user optimal cache placement processor and the unmanned aerial vehicle optimal cache placement processor, obtains the optimal cooperative cache placement information of the unmanned aerial vehicle and the user equipment in the current iteration process, judges whether the iteration times are smaller than the maximum iteration times, inputs the optimal user cache placement information into the unmanned aerial vehicle optimal cache placement processor if the iteration times are smaller than the maximum iteration times, and starts a new iteration; otherwise, outputting the current best unmanned aerial vehicle and user cooperative cache placement information, and ending the processing process.
The device for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network comprises an unmanned aerial vehicle optimal cache placing processor and a cache storing module, wherein the unmanned aerial vehicle optimal cache placing processor executes the following steps:
initializing cache placement information of the unmanned aerial vehicle k in a random mode;
under any iteration time t, updating each cache variable x of the unmanned aerial vehicle k according to the formula (1) by using a gradient descent algorithmt k,mTo obtain xt+1 k,m：
Figure BDA0002437006550000071
Wherein, L is the extended Lagrange function of the content acquisition time delay D of the network users, D is the content acquisition time delay of the network users,
Figure BDA0002437006550000072
is an arbitrary unmanned aerial vehicle,
Figure BDA0002437006550000073
is a set of unmanned aerial vehicles,
Figure BDA0002437006550000074
in order for any user to be able to do so,
Figure BDA0002437006550000075
in order to be a set of users,
Figure BDA0002437006550000076
in the case of any content, the content is,
Figure BDA0002437006550000077
is a collection of contents, xk,mCaching variables for unmanned aerial vehicles, xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0, matrix
Figure BDA0002437006550000078
Express all unmanned aerial vehicle's buffer memory and place information, unmanned aerial vehicle k's buffer memory is Q for spacekIndividual content representation, λkLagrange multiplier for drone k;
Figure BDA0002437006550000079
wherein λ iskBeing the lagrangian multiplier for drone k,
Figure BDA00024370065500000710
for the penalty term, ρ is a constant parameter for adjusting the convergence rate of the alternative direction multiplier method, and fn,mRepresenting the preference of user n for content m, S representing the size of the content, b being a ground base station, Rn,n′,Rk,n,Rb,nRespectively, the transmission rate of the wireless link between the user n and the adjacent user n', the unmanned aerial vehicle k and the base station B, Bk,nThe wireless backhaul link transmission rate assigned to user n for drone k,
Figure BDA00024370065500000711
for user access indication u n,n′1 means that user n establishes a D2D communication link with a neighboring user n' to obtain content, whereas un,n′＝0，uk,n1 means that the user n accesses the unmanned plane k to acquire the content, otherwise uk,n＝0，ub,n1 indicates that user n accesses to the ground base station b to obtain the content, otherwise ub,n＝0，yn',mCaching variables for users, y n',m1 denotes that user n' caches content m, whereas yn',m＝0；
According to the formula
Figure BDA0002437006550000081
All cache variables xk,m Value fixingBetween 0 and 1;
according to the formula
Figure BDA0002437006550000082
Updating Lagrange multiplier lambda corresponding to unmanned aerial vehicle kk；
Sequentially updating the cache variable, fixing the range of the cache variable and updating the Lagrange multiplier, and continuously iterating to judge whether the user content acquisition delay value after each iteration period is not changed any more than the last iteration period, namely | D (X)t+1 k)-D(Xt k)|<ε, ε is a small constant parameter, where D (X)t+1 k) Representing the user content acquisition delay value, D (X), after the current iteration cyclet k) Representing the time delay value of the previous iteration period;
the user time delay is still continuously reduced, and the iteration times are increased by 1;
when the user time delay is not changed, iteration is finished, relaxation is removed under the limitation of the cache space of the unmanned aerial vehicle, the cache variable value between 0 and 1 is converted into 0 or 1, and the optimal cache content X of the unmanned aerial vehicle k is obtainedk。
Unmanned aerial vehicle and user cooperate cache device of placing among unmanned aerial vehicle communication network, wherein optimize every unmanned aerial vehicle and cache variable xt+1 k,mThe gradient descent algorithm of (2) comprises the steps of:
calculating the gradient of the augmented Lagrangian function in each gradient descent iteration period p
Figure BDA0002437006550000083
According to the formula
Figure BDA0002437006550000084
Updating xk,mL is the learning rate;
according to formula Ip+1＝lp(0.96p) Updating the learning rate;
continuously iterating the three steps of calculating gradient, updating cache variable and updating learning rate, and judging that the value of the function L after each iteration period is higher than that of the last iteration periodWhether or not it is no longer changing, i.e. | L (x)p+1 k,m)-L(xp k,m)|<Epsilon, epsilon is a small constant parameter;
the value of the Lagrangian function L is still continuously reduced, and the iteration times are increased by 1;
the value of the Lagrangian function L is not changed, the iteration is finished, and the currently updated cache variable x is outputt+1 k,mThe value of (c).
The device for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network comprises a user optimal cache placing processor and a cache storing module, wherein the user optimal cache placing processor executes the following steps:
in any iteration cycle s of the greedy algorithm, calculating the content acquisition time delay delta D (delta D) of all network users, which can reduce the content of each user cache, and D (y)n,m＝1)-D(yn,m0), wherein, yn,mCaching variables for users, y n,m1 denotes that user n caches content m, whereas yn,m0, matrix
Figure BDA0002437006550000091
Cache placement information representing all users;
and comparing and finding out a user content pair with the largest delay reduction degree: { n, m } ═ max Δ D;
having the user store the content: y isn,m＝1；
Repeating the above steps, and judging whether the cache spaces of all the users are full after each iteration cycle, namely
Figure BDA0002437006550000092
Wherein QnRepresenting a cache space of user n;
if the user can still cache, adding 1 to the iteration times;
and (4) when all the user cache spaces are full, finishing iteration and outputting the optimal user cache placement information.
In the device for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network, the optimization control processor judges whether the current iteration period is less than the maximum iteration number, and if the current iteration period is less than the maximum iteration number, the user is cached and placed with confidenceTaking the information as input to carry out the next iteration, otherwise, finishing the iteration, and outputting the final unmanned aerial vehicle and user cooperative cache placement information { X }*,Y*}。
According to the technical scheme, the method and the device for cache placement of the unmanned aerial vehicle and the user system jointly consider the cache placement decision of the unmanned aerial vehicle and the user, so that the unmanned aerial vehicle and the user are mutually matched and cooperate with cache to obtain better system delay performance; meanwhile, for the problem of cooperative caching of the unmanned aerial vehicle and the user, a convergent optimization result is obtained in an iterative mode, and the method can obtain better time delay optimization performance under the condition of lower algorithm complexity.
Compared with the prior art, the invention has the advantages that: (1) the cache idea is introduced into the unmanned aerial vehicle communication network, and by deploying caches on the unmanned aerial vehicle and user equipment at the same time, content required by a user can be acquired from a neighboring user through D2D communication, or is directly provided by the unmanned aerial vehicle cache without requesting a core network with a longer distance, so that the time delay for the user to acquire the content can be effectively reduced, simultaneously, the resource and energy consumption caused by the wireless backhaul and downlink data transmission of the unmanned aerial vehicle are reduced, the system performance is improved, compared with the traditional mode of deploying caches at a certain position separately, the unmanned aerial vehicle and the user can be matched with each other, the content acquisition requirements of the user under different scenes are met, and the content acquisition time delay of the user is reduced together; (2) the content cached by the unmanned aerial vehicle and the user is optimized in a combined mode, the cooperative caching of the unmanned aerial vehicle and the user is carried out in an iterative mode, the user selects the content cached by the user on the basis of the caching decision of the unmanned aerial vehicle, the unmanned aerial vehicle adjusts the caching decision of the unmanned aerial vehicle on the basis of the caching contents of all the users, and compared with a method which is considered and optimized independently, the method can obtain the integral performance gain and further optimize the time delay performance of the system.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a schematic flow chart of a method for placing a cooperative cache between an unmanned aerial vehicle and a user in an unmanned aerial vehicle communication network according to an embodiment of the present invention;
fig. 2 is a schematic flow diagram illustrating placement of a cache of an unmanned aerial vehicle in an unmanned aerial vehicle communication network according to an embodiment of the present invention;
fig. 3 is a schematic flow chart of sequentially updating the cache variables of the unmanned aerial vehicle by using a gradient descent algorithm according to the embodiment of the present invention;
fig. 4 is a schematic flow chart illustrating user cache placement in an unmanned aerial vehicle communication network according to an embodiment of the present invention;
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as being fixedly connected, detachably connected, or integrally connected; may be a mechanical connection; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
In addition, in the description of the present invention, "a plurality" means two or more unless otherwise specified. In the present specification, the term "step" is used to mean not only an independent step but also a step that is not clearly distinguished from other steps, provided that the action intended by the step is achieved. In addition, the numerical range represented by "-" in the present specification means a range in which numerical values recited before and after "-" are included as a minimum value and a maximum value, respectively. In the drawings, elements having similar or identical structures are denoted by the same reference numerals.
Example one
Aiming at the problem of repeated transmission of a large amount of data in the Internet, in an unmanned aerial vehicle communication network, a caching technology can put some popular contents on edge nodes such as unmanned aerial vehicles and users which are closer to the users in advance, so that the users can acquire the contents more quickly and conveniently. The invention aims to provide a method for cooperatively caching and placing an unmanned aerial vehicle and a user in order to effectively reduce the content acquisition delay of the user in the whole network. For a communication network consisting of an unmanned aerial vehicle, a user and a ground base station, the unmanned aerial vehicle and the user have caching capacity at the same time, the caching decision of the user is guided by the caching content of the unmanned aerial vehicle through the combined optimization of the caching of the unmanned aerial vehicle and the user, and meanwhile, the caching content of the unmanned aerial vehicle is adjusted according to the caching decision of the user, so that the optimal storage content of caching nodes at different positions based on the whole network environment can be obtained. The method realizes the cooperative caching of the unmanned aerial vehicle and the user, can integrally improve the caching performance of the unmanned aerial vehicle communication network, and effectively reduces the time delay for the user to acquire the content.
Scene assumption is as follows: in the unmanned aerial vehicle communication network, there are b, K unmanned aerial vehicles of a ground base station
Figure BDA0002437006550000121
And N users
Figure BDA0002437006550000122
There are M contents in the network
Figure BDA0002437006550000123
Assuming that all the contents have the same size and are S bits, wherein any unmanned aerial vehicle is represented by k; any user is denoted by n; arbitrary content is denoted by m. The unmanned aerial vehicle and the user have caching capacity and use the matrix
Figure BDA0002437006550000124
Representing content cached by the drone, where xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0; by means of matrices
Figure BDA0002437006550000125
Representing the content of the user's cache, where y n,m1 denotes that user n caches content m, whereas yn,m0. Q for cache space of unmanned aerial vehicle kkContent representation, user n cache space QnA content representation. When a user requests for content, the user firstly checks the self cache, and if the content is cached, the content is directly obtained through the self cache; if the content is not cached, broadcasting a content request to other users in the communication range, and if the content is cached by nearby users, establishing a D2D communication link with the users for obtaining; if no nearby user caches the content, the user accesses the unmanned aerial vehicle or the ground base station to obtain the content according to the maximum signal-to-noise ratio, and at the moment, if the accessed unmanned aerial vehicle caches the content, the content copy stored by the unmanned aerial vehicle directly meets the content request; if the content is not cached by the unmanned aerial vehicle, the content needs to be requested to the core network through the backhaul link and then transmitted to the user. Therefore, the user can obtain the content through the self cache, the adjacent user cache, the unmanned aerial vehicle cache and the connected unmanned aerial vehicle or the base station through the core network request, and use un,iRepresenting the content service access situation of the user, i represents the object that the user can access to obtain the content, and according to the analysis, the content service access situation can be adjacent users, unmanned planes and ground base stations in the communication range of the user, namely
Figure BDA0002437006550000126
u n,i1 represents that user n establishes a communication link with i, whereas un,i0. Assuming that the base station is accessed to the core network through the high-capacity wired optical fiber line, and the data transmission delay between the base station and the core network is negligible, the delay for the user to acquire the content is:
Figure BDA0002437006550000131
wherein R isn,n′,Rk,n,Rb,nRespectively, the transmission rate of the wireless link between the user n and the adjacent user n', the unmanned aerial vehicle k and the base station B, Bk,nWireless backhaul link transmission rate, u, allocated to user n for drone kn,n,un,n′,uk,n,ub,nRespectively representing that a user n acquires contents through self cache, establishing a D2D communication link with an adjacent user n', accessing an unmanned aerial vehicle k and accessing a base station b, and y n',m1 denotes the user n' caches the content m, and S denotes the size of the content. User's request is influenced by content preferences, using fn,mRepresents the preference of the user n for the content m, 0 ≦ fn,m≤1，fn,mThe larger the preference of the user n to the content m is, the higher the probability of requesting the content m is, and therefore, the time delay for acquiring the content of the user in the whole network is as follows:
Figure BDA0002437006550000132
fig. 1 shows a flowchart of a method for cooperative cache placement of a drone and a user in a drone communication network, and in conjunction with fig. 1, the method includes the following steps:
the general flow of the principle used in this embodiment is: obtaining optimal unmanned aerial vehicle cache placement information X by using an ADMM (alternating direction multiplier method) algorithm according to the content cached by the current user equipment, and taking the unmanned aerial vehicle cache placement information X as input; obtaining optimal cache placement information Y of a user by using a global greedy algorithm according to the input cache placement information of the unmanned aerial vehicle; and taking the obtained unmanned aerial vehicle and user cache placement information as input of a new iteration to obtain the optimal unmanned aerial vehicle and user cache placement information in the next iteration period, repeating the iteration until the maximum iteration times are reached to obtain the final unmanned aerial vehicle and user cooperative cache placement information, wherein the unmanned aerial vehicle cache placement and the user cache placement are respectively executed once and recorded as an iteration period.
Step 101: starting the process;
step 102: obtaining cache placement information of the unmanned aerial vehicle through an ADMM (alternating direction multiplier method) algorithm;
FIG. 2 showsAnd acquiring a flow chart of the cache placement information of the unmanned aerial vehicle in the unmanned aerial vehicle communication network. According to the input user cache placement information, the content cache of each unmanned aerial vehicle is optimized in a distributed mode by using an ADMM (alternating direction multiplier method) algorithm to obtain the optimal unmanned aerial vehicle cache placement information, and the method further comprises the following steps: taking unmanned plane k as an example, first, content cache of unmanned plane k is initialized in a random manner
Figure BDA0002437006550000133
Then optimizing all buffer variables x of unmanned plane k through ADMM (alternating direction multiplier method) iteration processk,mFinally under the k buffer space limitation of the unmanned plane
Figure BDA0002437006550000141
And performing relaxation removal, and fixing the value of the cache variable to be 0 or 1. The specific steps of step 102 are described in detail below in conjunction with fig. 2.
Step 201: initializing cache placement information of the unmanned aerial vehicle k in a random mode;
step 202: under any iteration time t, updating each cache variable x of the unmanned aerial vehicle k according to the formula (1) by using a gradient descent algorithmt k,mTo obtain xt+1 k,m：
Figure BDA0002437006550000142
Wherein, L is the augmented Lagrange function of the content acquisition delay D of the network users, and D is the content acquisition delay of the network users:
Figure BDA0002437006550000143
wherein λ iskBeing the lagrangian multiplier for drone k,
Figure BDA0002437006550000144
rho is a penalty term and is a constant parameter for adjusting the convergence speed of the alternative direction multiplier method;
step 203: according to the formula(2) All cache variables xt+1 k,mThe value is fixed between 0 and 1:
Figure BDA0002437006550000145
step 204: according to the formula (3), updating the Lagrange multiplier lambda corresponding to the unmanned aerial vehicle kt kTo obtain lambdat+1 k：
Figure BDA0002437006550000151
Step 205: sequentially updating the cache variable, fixing the range of the cache variable and updating the Lagrange multiplier, and continuously iterating to judge whether the user content acquisition delay value after each iteration period is not changed any more than the last iteration period, namely | D (X)t+1 k)-D(Xt k)|<ε, ε is a small constant parameter, where D (X)t+1 k) Representing the user content acquisition delay value, D (X), after the current iteration cyclet k) Representing the time delay value of the previous iteration period; if the user delay D is changing, go to step 206; otherwise, go to step 207;
step 206: adding 1 to the iteration times;
step 207: after iteration is finished, loosening is carried out under the limitation of the cache space of the unmanned aerial vehicle, cache variable values between 0 and 1 are converted into 0 or 1, and the optimal cache content X of the k of the unmanned aerial vehicle is obtainedk。
And outputting the optimal unmanned aerial vehicle cache placement information X after all unmanned aerial vehicles operate the processes.
Fig. 3 shows a flow chart for sequentially updating the drone cache variables in step 202 using a gradient descent algorithm. Referring now to FIG. 3, step 202 is described in further detail:
step 301: calculating the cache variable x of the augmented Lagrange function under the condition of any gradient descent iteration times pp k,mGradient (2):
Figure BDA0002437006550000152
wherein
Figure BDA0002437006550000153
λkIn order to be a lagrange multiplier,
Figure BDA0002437006550000154
is a cache variable;
step 302: updating the cache variable according to equation (4):
Figure BDA0002437006550000155
wherein p is iteration times, and l is learning rate;
step 303: update the learning rate lp+1＝lp(0.96p)；
Step 304: continuously iterating the three steps of calculating gradient, updating cache variable and updating learning rate, and judging whether the value of the function L after each iteration number is not changed any more than the last iteration period, namely | L (x)p+1 k,m)-L(xp k,m)|<Epsilon, epsilon is a small constant parameter; if the lagrangian function value L is changing, go to step 305; otherwise, go to step 306;
step 305: adding 1 to the iteration times;
step 306: after the iteration is finished, outputting the currently updated cache variable xt+1 k,mThe value of (c).
Step 103: obtaining user cache placement information through a global greedy algorithm;
fig. 4 shows a flow chart of user cache placement in a drone communication network. After the optimal cache placement information of the unmanned aerial vehicle is obtained, the optimal user cache placement information in the current state is obtained by using a global greedy algorithm, and step 103 is further described in detail below with reference to fig. 4.
Step 401: in any iteration cycle s of the greedy algorithm, calculating the content acquisition time delay delta D (delta D) of all network users, which can reduce the content of each user cache, and D (y)n,m＝1)-D(yn,m＝0)；
Step 402: and comparing and finding out a user content pair with the largest delay reduction degree: { n, m } ═ max Δ D;
step 403: having the user store the content: y isn,m＝1；
Step 404: the above steps are iterated continuously, and whether the cache spaces of all the users are full after each iteration period is judged, namely that
Figure BDA0002437006550000161
Wherein QnRepresenting a cache space of user n; if yes, go to step 406; otherwise, go to step 405;
step 405: adding 1 to the iteration times;
step 406: and (5) finishing the iteration and outputting the optimal user cache placement information Y.
Step 104: judging whether the iteration times r of the unmanned aerial vehicle cache placement and the user cache placement reach the maximum iteration times; if yes, go to step 106; otherwise, step 105 is performed.
Step 105: adding 1 to the iteration number, namely r is r +1 and returning to the step 102;
step 106: outputting the cache placement information of the unmanned aerial vehicle and the user at the moment;
step 107: the flow ends.
Through continuous iteration steps 102 and 103 until the maximum iteration times are reached, the optimal unmanned aerial vehicle and user cooperative cache placement information is obtained, and the system delay performance is greatly improved.
Example two
The embodiment of the invention also discloses a device for cooperatively caching and placing the unmanned aerial vehicle and the user in the unmanned aerial vehicle communication network. The structural composition of the unmanned aerial vehicle and user cooperative cache placement device is briefly introduced below, and the device is deployed on a macro base station to optimize the content caches of all unmanned aerial vehicles and users in the network. The device includes: the system comprises an unmanned aerial vehicle optimal cache placement processor, a user optimal cache placement processor and an optimization control processor.
The optimal cache placement processor of the unmanned aerial vehicle is connected with the optimal cache placement processor of the user and the optimization control processor. According to the user cache placement information input by the optimization control processor, the optimal unmanned aerial vehicle cache placement information at the moment is obtained by using an ADMM (alternating direction multiplier method), and the information is input into the user optimal cache placement processor.
The user optimal cache placement processor is connected with the unmanned aerial vehicle optimal cache placement processor and the optimization control processor. And obtaining the optimal user cache placement information at the moment by using a global greedy algorithm according to the unmanned aerial vehicle cache placement information input by the unmanned aerial vehicle optimal cache placement processor, and inputting the information into the optimization control processor.
The optimal control processor is connected with the user optimal cache placement processor and the unmanned aerial vehicle optimal cache placement processor. The optimization control processor obtains the optimal cooperative cache placement information of the unmanned aerial vehicle and the user equipment in the current iteration process, judges whether the iteration times are smaller than the maximum iteration times, and inputs the optimal user cache placement information into the optimal cache placement processor of the unmanned aerial vehicle to start a new iteration if the iteration times are smaller than the maximum iteration times; otherwise, outputting the current best unmanned aerial vehicle and user cooperative cache placement information, and ending the processing process.
The following specifically introduces the operation process of the unmanned aerial vehicle communication network unmanned aerial vehicle and the user cooperative cache placement device, specifically:
in a communication network consisting of a plurality of drones, a set of drones
Figure BDA0002437006550000171
Any unmanned aerial vehicle is denoted by k; user collection
Figure BDA0002437006550000172
Any user is denoted by n; content aggregation in a network
Figure BDA0002437006550000173
Arbitrary content is denoted by m. Unmanned aerial vehicles and users in the system are provided with cache devices. By means of matrices
Figure BDA0002437006550000174
Representing content cached by the drone, where xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0. By means of matrices
Figure BDA0002437006550000181
Representing the content of the user's cache, where y n,m1 denotes that user n caches content m, whereas yn,m0. Q for cache space of unmanned aerial vehicle kkContent representation, user n cache space QnA content representation.
In an optimal cache placement processor of the unmanned aerial vehicle, obtaining optimal cache content X of the unmanned aerial vehicle by using an ADMM (alternating direction multiplier method) algorithm according to input user cache placement information; in a user optimal cache placement processor, obtaining user optimal cache content Y by using a global greedy algorithm according to input unmanned aerial vehicle cache placement information; in the optimization control processor, the obtained unmanned aerial vehicle and user cache placement information is used as the input of a new iteration to obtain the best content of the unmanned aerial vehicle and the user cache in the next iteration period, the iteration is repeated until the preset iteration times are reached, and the unmanned aerial vehicle and user cooperative cache placement information at the moment is output. Specific treatment processes are described in detail below.
In the processor for placing the optimal cache of the unmanned aerial vehicle, according to the input user cache placement information, the content cache of each unmanned aerial vehicle is optimized in a distributed manner by using an ADMM (alternating direction multiplier method) algorithm to obtain the optimal cache placement information of the unmanned aerial vehicle, and the method further comprises the following steps: taking unmanned plane k as an example, first, content cache of unmanned plane k is initialized in a random manner
Figure BDA0002437006550000182
Then optimizing all buffer variables x of unmanned plane k through ADMM (alternating direction multiplier method) iteration processk,mFinally under the k buffer space limitation of the unmanned plane
Figure BDA0002437006550000183
And performing relaxation removal, and fixing the value of the cache variable to be 0 or 1. In particular, the amount of the solvent to be used,in the iterative process of ADMM (alternating direction multiplier method), under any iteration number t, firstly, a gradient descent algorithm is used for expressing the formula
Figure BDA0002437006550000184
Sequentially optimizing each cache variable of unmanned aerial vehicle k to obtain xt+1 k,mWherein, L is the augmented Lagrange function of the content acquisition delay D of the network users, and D is the content acquisition delay of the network users:
Figure BDA0002437006550000191
wherein λ iskThe method is characterized in that the method is a Lagrange multiplier of the unmanned aerial vehicle k, and rho is a constant parameter for adjusting the convergence speed of the ADMM (alternating direction multiplier method); then according to formula
Figure BDA0002437006550000192
X obtainedt+1 k,mThe value is limited to 0-1; finally according to formula
Figure BDA0002437006550000193
Updating lagrange multiplier lambdat+1 k. The above process is iterated until the content acquisition delay of the network-wide user is not changed, i.e. | D (X)t+1 k)-D(Xt k)|<Epsilon and epsilon are smaller constant parameters, and the iteration is finished to obtain the optimal cache content X of the current unmanned aerial vehicle kk. The other unmanned aerial vehicles also process the same, and finally the best unmanned aerial vehicle cache placement information X is obtained.
Wherein optimizing each drone cache variable xt+1 k,mThe gradient descent algorithm comprises the following steps: in each gradient descent iteration number p, the gradient of the augmented Lagrangian function is calculated firstly
Figure BDA0002437006550000194
Then according to formula
Figure BDA0002437006550000195
Updating xk,mL is the learning rate; finally according to formula lp+1＝lp(0.96p) The learning rate is updated. The above process is iterated until the value of the function L no longer changes, i.e. | L (x)p+1 k,m)-L(xp k,m)|<ε, the iteration ends, gets the updated x at this timet+1 k,m。
In the user optimal cache placement processor, according to the input unmanned aerial vehicle cache placement information, the optimal user cache placement information under the current state is obtained by using a global greedy algorithm, and the method further comprises the following steps: in any iteration cycle s of the greedy algorithm, calculating the content acquisition delay of the full-network users, namely delta D (delta D) D (y), of which the cache content m of the user n can be reducedn,m＝1)-D(yn,m0) and then find one of them, whose delay is reduced most, by comparison, let its y ben,m1 is ═ 1; the above process is repeated until all user cache spaces are full, i.e. all user cache spaces are full
Figure BDA0002437006550000201
And finishing the iteration to obtain the optimal user cache placement information Y at the moment.
So far, the optimal contents { X ] cooperatively cached by the unmanned aerial vehicle and the user in a certain integral iteration period r are obtainedr,Yr}. The optimization control processor judges whether the current iteration period is less than the maximum iteration number, if so, the user cache placement information is used as input for next iteration, otherwise, the iteration is finished, and the final unmanned aerial vehicle and user cooperative cache placement information { X at the moment is output*,Y*}。
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solution of the present invention by those skilled in the art should fall within the protection scope defined by the claims of the present invention without departing from the spirit of the present invention.

Claims (8)

1. A method for cooperatively caching and placing an unmanned aerial vehicle and a user in an unmanned aerial vehicle communication network comprises the following steps:
obtaining cache placement information of the unmanned aerial vehicle by an alternating direction multiplier method, and obtaining user cache placement information by a global greedy algorithm;
continuously iterating the cache placement information of the unmanned aerial vehicle and the cache placement information of the user, and judging whether the maximum iteration times is reached;
if the iteration does not reach the maximum number of times, adding 1 to the iteration number, and iterating the cache placement information of the unmanned aerial vehicle and the cache placement information of the user again;
if the iteration reaches the maximum times, outputting the cache placement information of the unmanned aerial vehicle and the user at the moment;
the method for obtaining the cache placement information of the unmanned aerial vehicle by the alternating direction multiplier method further comprises the following steps:
step 201: initializing cache placement information of the unmanned aerial vehicle k in a random mode;
step 202: any iteration number t is calculated according to a formula by using a gradient descent algorithm
Figure FDA0003010306180000011
Sequentially updating each cache variable x of unmanned aerial vehicle kt k,mTo obtain xt+1 k,m；
Wherein L is an augmented Lagrange function of the content acquisition delay D of the whole network user,
Figure FDA0003010306180000012
is an arbitrary unmanned aerial vehicle,
Figure FDA0003010306180000013
is a set of unmanned aerial vehicles,
Figure FDA0003010306180000014
in order for any user to be able to do so,
Figure FDA0003010306180000015
in order to be a set of users,
Figure FDA0003010306180000016
in the case of any content, the content is,
Figure FDA0003010306180000017
is a collection of contents, xk,mCaching variables for unmanned aerial vehicles, xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0, matrix
Figure FDA0003010306180000021
Express all unmanned aerial vehicle's buffer memory and place information, unmanned aerial vehicle k's buffer memory is Q for spacekIndividual content representation, λkLagrange multiplier for drone k;
step 203: according to the formula
Figure FDA0003010306180000022
All cache variables xt+1 k,mThe value is fixed between 0 and 1;
step 204: according to the formula
Figure FDA0003010306180000023
Updating Lagrange multiplier lambda corresponding to unmanned aerial vehicle kk：λt kTo obtain lambdat+1 kRho is a constant parameter for adjusting the convergence speed of the alternative direction multiplier method;
step 205: sequentially updating the cache variable, fixing the range of the cache variable and updating the Lagrange multiplier, and continuously iterating to judge whether the user content acquisition delay value after each iteration period is not changed any more than the last iteration period, namely | D (X)t +1 k)-D(Xt k)|<ε, ε is a small constant parameter, where D (X)t+1 k) Representing the user content acquisition delay value, D (X), after the current iteration cyclet k) Representing the time delay value of the previous iteration period; if the user delay D is changing, go to step 206; otherwise, go to step 207;
step 206: adding 1 to the iteration times;
step 207: the iteration is finished under the limitation of the cache space of the unmanned aerial vehiclePerforming relaxation removal, converting the cache variable value between 0 and 1 into 0 or 1, and obtaining the optimal cache content X of the unmanned aerial vehicle kk。
2. The method of claim 1, wherein L is an augmented lagrangian function of a network-wide user content acquisition delay D, and D is a network-wide user content acquisition delay:
Figure FDA0003010306180000031
Figure FDA0003010306180000032
wherein λ iskBeing the lagrangian multiplier for drone k,
Figure FDA0003010306180000033
for the penalty term, ρ is a constant parameter for adjusting the convergence rate of the alternative direction multiplier method, and fn,mRepresenting the preference of user n for content m, S representing the size of the content, b being a ground base station, Rn,n′,Rk,n,Rb,nRespectively, the transmission rate of the wireless link between the user n and the adjacent user n', the unmanned aerial vehicle k and the base station B, Bk,nThe wireless backhaul link transmission rate assigned to user n for drone k,
Figure FDA0003010306180000034
for user access indication un,n′1 means that user n establishes a D2D communication link with a neighboring user n' to obtain content, whereas un,n′＝0，uk,n1 means that the user n accesses the unmanned plane k to acquire the content, otherwise uk,n＝0，ub,n1 indicates that user n accesses to the ground base station b to obtain the content, otherwise ub,n＝0，yn',mCaching variables for users, yn',m1 denotes that user n' caches content m, whereas yn',m＝0。
3. The method for cooperative cache placement of drones and users in a drone communication network according to claim 2, wherein step 202 further comprises the steps of:
step 301: calculating the cache variable x of the augmented Lagrange function under the condition of any gradient descent iteration times pp k,mGradient (2):
Figure FDA0003010306180000035
wherein
Figure FDA0003010306180000036
λkIn order to be a lagrange multiplier,
Figure FDA0003010306180000037
is a cache variable;
step 302: according to the formula xp+1 k,m＝xp k,m-lp▽L(xp k,m) Updating a cache variable, wherein p is iteration times and l is a learning rate;
step 303: update the learning rate lp+1＝lp(0.96p)；
Step 304: continuously iterating the three steps of calculating gradient, updating cache variable and updating learning rate, and judging whether the value of the function L after each iteration period is not changed any more than the last iteration period, namely | L (x)p+1 k,m)-L(xp k,m)|<Epsilon, epsilon is a small constant parameter; if the lagrangian function value L is changing, go to step 305; otherwise, go to step 306;
step 305: adding 1 to the iteration times;
step 306: after the iteration is finished, outputting the currently updated cache variable xt+1 k,mThe value of (c).
4. The method for collaborative cache placement by drone and user in a drone communication network of claim 3, wherein obtaining user cache placement information by a global greedy algorithm further comprises the steps of:
step 401: in any iteration period of the greedy algorithm, calculating the content acquisition time delay delta D (delta D) of all network users, which can reduce the content of each user cache, and D (y)n,m＝1)-D(yn,m0), wherein, yn,mCaching variables for users, yn,m1 denotes that user n caches content m, whereas yn,m0, matrix
Figure FDA0003010306180000041
Cache placement information representing all users;
step 402: and comparing and finding out a user content pair with the largest delay reduction degree: { n, m } ═ max Δ D;
step 403: having the user store the content: y isn,m＝1；
Step 404: the above steps are iterated continuously, and whether the cache spaces of all the users are full after each iteration period is judged, namely that
Figure FDA0003010306180000042
Wherein QnRepresenting a cache space of user n; if yes, go to step 406; otherwise, go to step 405;
step 405: adding 1 to the iteration times;
step 406: and (5) finishing the iteration and outputting the optimal user cache placement information Y.
5. A device for cooperative cache placement of an unmanned aerial vehicle and a user in an unmanned aerial vehicle communication network is deployed on a macro base station and comprises: an unmanned aerial vehicle optimal cache placement processor, a user optimal cache placement processor and an optimization control processor; wherein,
the optimal cache placement processor of the unmanned aerial vehicle is connected with the optimal cache placement processor of the user and the optimization control processor, and the optimal cache placement information of the unmanned aerial vehicle at the moment is obtained by utilizing an alternating direction multiplier method according to the user cache placement information input by the optimization control processor and is input into the optimal cache placement processor of the user;
the user optimal cache placement processor is connected with the unmanned aerial vehicle optimal cache placement processor and the optimization control processor, and according to unmanned aerial vehicle cache placement information input by the unmanned aerial vehicle optimal cache placement processor, the optimal user cache placement information at the moment is obtained by using a global greedy algorithm and is input into the optimization control processor;
the optimization control processor is connected with the user optimal cache placement processor and the unmanned aerial vehicle optimal cache placement processor, obtains the optimal cooperative cache placement information of the unmanned aerial vehicle and the user equipment in the current iteration process, judges whether the iteration times are smaller than the maximum iteration times, inputs the optimal user cache placement information into the unmanned aerial vehicle optimal cache placement processor if the iteration times are smaller than the maximum iteration times, and starts a new iteration; otherwise, outputting the current best unmanned aerial vehicle and user cooperative cache placement information, and ending the processing process;
the optimal cache placement processor of the unmanned aerial vehicle executes the following steps:
initializing cache placement information of the unmanned aerial vehicle k in a random mode;
under any iteration time t, utilizing a gradient descent algorithm according to a formula
Figure FDA0003010306180000051
Sequentially updating each cache variable x of unmanned aerial vehicle kt k,mTo obtain xt+1 k,m；
Wherein, L is the extended Lagrange function of the content acquisition time delay D of the network users, D is the content acquisition time delay of the network users,
Figure FDA0003010306180000052
is an arbitrary unmanned aerial vehicle,
Figure FDA0003010306180000053
is a set of unmanned aerial vehicles,
Figure FDA0003010306180000061
in order for any user to be able to do so,
Figure FDA0003010306180000062
in order to be a set of users,
Figure FDA0003010306180000063
in the case of any content, the content is,
Figure FDA0003010306180000064
is a collection of contents, xk,mCaching variables for unmanned aerial vehicles, xk,mIndicate unmanned aerial vehicle k to buffer content m for 1, otherwise xk,m0, matrix
Figure FDA0003010306180000065
Express all unmanned aerial vehicle's buffer memory and place information, unmanned aerial vehicle k's buffer memory is Q for spacekIndividual content representation, λkLagrange multiplier for drone k;
Figure FDA0003010306180000066
Figure FDA0003010306180000067
wherein λ iskBeing the lagrangian multiplier for drone k,
Figure FDA0003010306180000068
for the penalty term, ρ is a constant parameter for adjusting the convergence rate of the alternative direction multiplier method, and fn,mRepresenting the preference of user n for content m, S representing the size of the content, b being a ground base station, Rn,n′,Rk,n,Rb,nRespectively, the transmission rate of the wireless link between the user n and the adjacent user n', the unmanned aerial vehicle k and the base station B, Bk,nThe wireless backhaul link transmission rate assigned to user n for drone k,
Figure FDA0003010306180000069
for user access indication un,n′1 means that user n establishes a D2D communication link with a neighboring user n' to obtain content, whereas un,n′＝0，uk,n1 means that the user n accesses the unmanned plane k to acquire the content, otherwise uk,n＝0，ub,n1 indicates that user n accesses to the ground base station b to obtain the content, otherwise ub,n＝0，yn',mCaching variables for users, yn',m1 denotes that user n' caches content m, whereas yn',m＝0；
According to the formula
Figure FDA00030103061800000610
All cache variables xk,mThe value is fixed between 0 and 1;
according to the formula
Figure FDA00030103061800000611
Updating Lagrange multiplier lambda corresponding to unmanned aerial vehicle kk；
Sequentially updating the cache variable, fixing the range of the cache variable and updating the Lagrange multiplier, and continuously iterating to judge whether the user content acquisition delay value after each iteration period is not changed any more than the last iteration period, namely | D (X)t+1 k)-D(Xt k)|<ε, ε is a small constant parameter, where D (X)t+1 k) Representing the user content acquisition delay value, D (X), after the current iteration cyclet k) Representing the time delay value of the previous iteration period;
the user time delay is still continuously reduced, and the iteration times are increased by 1;
when the user time delay is not changed, iteration is finished, relaxation is removed under the limitation of the cache space of the unmanned aerial vehicle, the cache variable value between 0 and 1 is converted into 0 or 1, and the optimal cache content X of the unmanned aerial vehicle k is obtainedk。
6. The apparatus of claim 5, wherein the optimization is based on the cooperative placement of the UAV and the user cache in the UAV communication networkEach unmanned aerial vehicle caches variable xt+1 k,mThe gradient descent algorithm of (2) comprises the steps of:
in each gradient descent iteration number p, a gradient ^ L (x) of the augmented Lagrangian function is calculatedp k,m)；
According to the formula xp+1 k,m＝xp k,m-lp▽L(xp k,m) Updating xk,mL is the learning rate;
according to formula Ip+1＝lp(0.96p) Updating the learning rate;
continuously iterating the three steps of calculating gradient, updating cache variable and updating learning rate, and judging whether the value of the function L after each iteration period is not changed any more than the last iteration period, namely | L (x)p+1 k,m)-L(xp k,m)|<Epsilon, epsilon is a small constant parameter;
the value of the Lagrangian function L is still continuously reduced, and the iteration times are increased by 1;
the value of the Lagrangian function L is not changed, the iteration is finished, and the currently updated cache variable x is outputt+1 k,mThe value of (c).
7. The apparatus of claim 6, wherein the processor performs the following steps:
in any iteration period of the greedy algorithm, calculating the content acquisition time delay delta D (delta D) of all network users, which can reduce the content of each user cache, and D (y)n,m＝1)-D(yn,m0), wherein, yn,mCaching variables for users, yn,m1 denotes that user n caches content m, whereas yn,m0, matrix
Figure FDA0003010306180000081
Cache placement information representing all users;
and comparing and finding out a user content pair with the largest delay reduction degree: { n, m } ═ max Δ D;
let the user store theThe content is as follows: y isn,m＝1；
Repeating the above steps, and judging whether the cache spaces of all the users are full after each iteration cycle, namely
Figure FDA0003010306180000082
Wherein QnRepresenting a cache space of user n;
if the user can still cache, adding 1 to the iteration times;
and (4) when all the user cache spaces are full, finishing the iteration and outputting the optimal user cache placement information Y.
8. The apparatus of claim 7, wherein the optimization control processor determines whether a current iteration cycle is less than a maximum number of iterations, and if so, takes the user cache placement information as input for the next iteration, otherwise, the iteration is terminated, and outputs final unmanned aerial vehicle and user collaborative cache placement information { X } at that time*,Y*}。
CN202010255135.8A 2020-04-02 2020-04-02 Unmanned aerial vehicle and user cooperative cache placement method and device Active CN111464231B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010255135.8A CN111464231B (en) 2020-04-02 2020-04-02 Unmanned aerial vehicle and user cooperative cache placement method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010255135.8A CN111464231B (en) 2020-04-02 2020-04-02 Unmanned aerial vehicle and user cooperative cache placement method and device

Publications (2)

Publication Number Publication Date
CN111464231A CN111464231A (en) 2020-07-28
CN111464231B true CN111464231B (en) 2021-06-22

Family

ID=71681163

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010255135.8A Active CN111464231B (en) 2020-04-02 2020-04-02 Unmanned aerial vehicle and user cooperative cache placement method and device

Country Status (1)

Country Link
CN (1) CN111464231B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111970046B (en) * 2020-08-18 2022-06-21 北京邮电大学 Cooperative transmission method for unmanned aerial vehicle network dynamic cache deployment
CN111988762B (en) * 2020-09-01 2022-03-29 重庆邮电大学 Energy efficiency maximum resource allocation method based on unmanned aerial vehicle D2D communication network
CN112260746B (en) * 2020-10-16 2022-01-07 北京邮电大学 Site arrangement, cache placement and resource allocation method and device
CN112702719A (en) * 2020-11-27 2021-04-23 北京工业大学 High-energy-efficiency D2D resource allocation method in unmanned aerial vehicle scene
CN113163377B (en) * 2021-04-25 2022-04-01 北京邮电大学 Unmanned aerial vehicle network deployment and resource allocation method and device
CN113411862B (en) * 2021-06-18 2022-08-09 北京邮电大学 Cache placement and user access method and device in dynamic cellular network
CN113891442B (en) * 2021-10-27 2023-07-07 哈尔滨工业大学（深圳） Unmanned aerial vehicle relay transmission energy efficiency optimization method based on D2D communication
CN114980139A (en) * 2022-05-25 2022-08-30 西安电子科技大学 Capacity coverage enhancement method for air-ground wireless network access and return integrated system
CN115021798B (en) * 2022-06-02 2023-08-01 北京邮电大学 Caching, communication and control method and system for multi-unmanned aerial vehicle network

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8838724B2 (en) * 2010-07-02 2014-09-16 Futurewei Technologies, Inc. Computation of caching policy based on content and network constraints
US8811614B2 (en) * 2012-10-16 2014-08-19 The Boeing Company Space based authentication utilizing signals from low and medium earth orbit
US9401868B2 (en) * 2012-11-08 2016-07-26 Futurewei Technologies, Inc. Method of traffic engineering for provisioning routing and storage in content-oriented networks
EP3139515A1 (en) * 2015-09-02 2017-03-08 BAE Systems PLC Apparatus and method for communications management
WO2017070635A1 (en) * 2015-10-22 2017-04-27 Phluido, Inc. Virtualization and orchestration of a radio access network
EP3756283A1 (en) * 2018-02-23 2020-12-30 Telefonaktiebolaget LM Ericsson (publ) Method and controller for controlling a wireless link
CN110417847B (en) * 2019-01-09 2020-09-01 北京邮电大学 Method and device for user access and content caching in unmanned aerial vehicle communication network

Also Published As

Publication number Publication date
CN111464231A (en) 2020-07-28

Similar Documents

Publication Publication Date Title
CN111464231B (en) Unmanned aerial vehicle and user cooperative cache placement method and device
CN110417847B (en) Method and device for user access and content caching in unmanned aerial vehicle communication network
CN109413724B (en) MEC-based task unloading and resource allocation scheme
CN110996293B (en) Network deployment and resource allocation method and system for unmanned aerial vehicle
CN109413615B (en) Energy latency tradeoff for MEC-based energy-aware offloading in the Internet of vehicles
CN109194763B (en) Caching method based on small base station self-organizing cooperation in ultra-dense network
CN108093435B (en) Cellular downlink network energy efficiency optimization system and method based on cached popular content
CN107889160A (en) A kind of cell network edge part caching method for considering user&#39;s time delay
CN113163377B (en) Unmanned aerial vehicle network deployment and resource allocation method and device
CN108259628B (en) Content caching and user association joint optimization method in heterogeneous cellular network
CN110290507A (en) A kind of cache policy and frequency spectrum distributing method of D2D communication assistant edge caching system
CN110602722B (en) Design method for joint content pushing and transmission based on NOMA
CN106792995B (en) User access method for guaranteeing low-delay content transmission in 5G network
CN115801091B (en) Large-scale constellation network resource scheduling method for satellite-ground cooperative computing
Khan et al. On the application of agglomerative hierarchical clustering for cache-assisted D2D networks
CN108521640A (en) A kind of content distribution method in cellular network
CN113993168B (en) Collaborative caching method based on multi-agent reinforcement learning in fog wireless access network
CN108965034B (en) Method for associating user to network under ultra-dense deployment of small cell base station
Jiang et al. Double DQN based computing offloading scheme for fog radio access networks
CN108882269A (en) The super-intensive network small station method of switching of binding cache technology
CN109547979B (en) Content distribution method combining energy consumption and user fairness under D2D cache network
CN115278779B (en) VR service module dynamic placement method based on rendering perception in MEC network
CN110933687A (en) User uplink and downlink access method and system based on decoupling
CN113411862B (en) Cache placement and user access method and device in dynamic cellular network
CN113301576B (en) Cellular network resource allocation method based on improved genetic algorithm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant